Question of Valency in Gaseous Ionization. 549 



charge *. In this paper it is shown that an oil-drop, even 

 when it already carries many negative charges, and therefore 

 repels negative ions, nevertheless more frequently takes up 

 negative than positive charges. This is the more striking, 

 since in a condenser, whether it be charged or not, the 

 number of the negative ions is smaller than that of the 

 positive, which is due to the greater coefficient of diffusion 

 and velocity of the negative ions. Even the increased 

 frequency of " catches " of negative ions is to be explained 

 by their greater diffusion ; for, according to Ebert |, this 

 adsorption of the ions is proportional to diffusion and 

 velocity. Thus, with an equal number of ions of both signs, 

 the ratio of " catches" for negative and positive ions should 

 be 1*36. The ratio of diffusion of the doubly-charged ions 

 to that of the singly-charged ones is 2 ; with those trebly- 

 charged, a ratio of 3 is to be expected ; and so on. Thus, 

 due to the diffusion only, the observed number of doubly- 

 charged ions must be multiplied by 2 ; and so on for the 

 higher-charged ions. And this result would be the same, 

 even if the diffusion of the multiply charged ions were the 

 same as that of the single ones, because then the velocity of 

 the variously charged ions would increase by the factors 2 

 and 3, and so on, so that the number of the corresponding 

 ions in the gas would become smaller in the same ratio. 

 Now let us consider Table II L in the paper of Millikan and 

 Fletcher. We take this table, since it is only in this series 

 that there are any observations which may possibly refer to 

 doubly-charged ions. This possibility is even expressed by 

 the authors themselves, but given up in favour of another 

 explanation. In 79 cases which were observed, 76 showed a 

 single charge, 3 a multiple charge. Of the latter, 2 are 

 marked out as dubious. We may, however, take 2 of these 

 observed 3 cases as real. If we take into account the 

 diffusion, these 2 cases indicate at least 5 per cent, of the 

 ions to be multiply-charged if there were only doubly-charged 

 ions, and even more if there should be greater charges. 

 Especially is this true if we consider that the multiply- 

 charged ones are repelled much more violently from the 

 positive oil-drop than the single ones. All these modifi- 

 cations of the ionic constants, the increased recombination, 

 and especially the smaller diffusion and the increased re- 

 pulsion from the drop, work together so as to cause the 

 double charges to appear rarer than they really are. Indeed, 



* R. A. Millikan, Phil. Mag. [&] vol. xix. p. 209 (1910). 

 t H. Ebert, Jakrb. d, Had, u.'Elektr. vol. iii. p. 61 (1906). 



Phil May. S. 6. Vol. 22. No. 130. Oct. 1911. 2 



